There are quite a number of medium format digital cameras available today; the vast majority are designed to handle like oversize DSLRs, and in some cases, there’s very little difference size- and control-wise between these cameras – take the Leica S, for instance. This makes them both familiar and easy to use, but also somewhat liable to catch out the unweary. My digital field work with medium format is done with a Hasselblad CFV-39, mounted on a 501CM body. The method of operation constantly reminds you that this is most certainly not another DSLR; not least because you have to wind the camera after every shot to recock the shutter and lower the mirror! The intention of this article is to look at the practicalities – or impracticalities – of using medium format digital in the field or while travelling as a DSLR replacement, and more importantly, in a way that lets you actually see enough of a difference to justify it in the first place.
Myth number one: it’s expensive.
It isn’t. Second hand digital backs are falling in price; I suspect we’re about to see another drop – especially on the current generation – once the new CMOS censored cameras hit the market in force; already Phase One, Hasselblad and Pentax have announced cameras around the Sony 50MP chip. I think there will still remain some die-hard CCD fans who won’t switch to CMOS, however – there are a number of reasons to still favour a CCD for some applications. Using a camera that’s one or two generations behind is not going to yield poor image quality; remember that these were the state of the art commercial tools at their release, and frankly, the bar hasn’t moved all that much in this market. Limited sales and limited R&D budgets have seen to that. So long as you are conscious of the limitations of the large CCDs, you can still produce excellent results.
Myth number two: you can’t use the in the dark.
This is partially true: don’t expect to use your Phase One as a reportage camera; firstly, you’re already at least a stop or two down on the lens front compared to a DSLR with f1.4 primes; there’s no stabilisation on any medium format camera, and finally, your sensor realistically tops out at ISO 400, perhaps ISO 800 at a push with some noise reduction. Honestly: the CFV-39 at ISO 800 looks like the D800E at ISO 6400. That says two things: just how clean the D800E is, and how large its shooting envelope; but also that you’re not constrained to ISO 50 on the digital back.
Myth number three: they’re power hungry.
If you’re powering a whole camera off it, perhaps; I don’t have experience with the Phase One cameras, but I do know that the H series ‘Blads can manage at least 400-500 shots per charge; the Leica S is even more – easily a thousand – and even the CFV-39 and -50 will manage ~700-800. Bear in mind that you’re not going to be shooting one of these things like a D4 – in single shot mode, with carefully considered setup and framing, you may realistically shoot a couple of hundred images a day at most.
My choice of medium format digital has perhaps the most restrictions of any of its kind:
- Manual focus only.
- The body is designed for waist level work, since the concept of portrait orientation doesn’t exist for a 6×6 square frame. This has viewing and ergonomic consequences, that can only be solved with a right-angle finder – the HC-3 or HC-4 – and a jury-rigged hand grip made from an L bracket (that’s also useful for mounting the camera to a tripod).
- You have to wind it between shots
- The leaf shutter may be silent and low vibration, but the enormous mirror and auxiliary curtains are most certainly not: remember that this camera, its optics, its viewfinder system, and all of the moving parts were designed for a 6×6 frame – at most, the digital backs use a 1.1x crop of 645. That’s a lot of extra moving mass.
- The camera body is fully mechanical, and thus has no battery. Interface between the body and back is by means of the little pin that normally advances the film counter; it tells the back to wake up and fire the sensor, which is powered on for a preset duration, during which time the exposure happens. Surprisingly, there is almost zero noticeable lag. In any case, it requires winding to recock between each image – which means single shot only.
- The back was released in 2009, with electronics that I think date back even earlier. Certainly the LCD, graphics and other UI elements feel like it; it’s not slow, but it’s not really snappy, either. The LCD is only useful for a) confirming that the camera has taken a shot, and b) viewing the histogram. It is not useful for anything else, and unlike modern DSLRs, cannot be used for doing a first cut evaluation of images.
- There’s no meter: it’s not only very, very intolerant to underexposure and subsequent recovery, but also moderately intolerant to overexposure. Use an external spot meter, recalibrate your eyeballs for higher accuracy, or be prepared for iterations using the histogram.
You may be wondering why I chose it. One answer is simple: because the process of operating it most closely matches the experience of shooting film, whose output I like very much; the only difference is that you are composing for 645 and not 6×6, so portrait orientation or a crop factor comes into play. Nevertheless, the experience feels very, very similar and familiar – except you do have instant gratification – though admittedly I almost always shoot it with the LCD off, and treat the back like it’s loaded with film – very unforgiving slide film, because one of the properties of a CCD is that dynamic range is great – but only if you get the exposure right at the time of capture. Pushing and pulling a CCD file is not generally a good idea because the shadows aren’t anywhere near as clean, nor are the highlight transitions smooth and easily recoverable.
Note grip and eye-level prism; in this configuration, the camera is bulky but handles like a DSLR – albeit one that you have to wind and whose shutter button you have to trigger with your right fourth finger.
Many things that need workarounds, the biggest one of which is stability. The CFV-39 requires 1/2x at the bare minimum, or 1/3x preferably – we’re not talking 1/2x the 35mm equivalent, we’re taking 1/2x the actual focal length to be used handheld. That means 1/125s minimum for the 80mm, and preferably 1/250s+ to ensure a sharp image – assuming of course you focused it on the money to begin with. There’s another problem: the leaf shutters max out at 1/500s, which means you don’t really have a lot of latitude. Using the prism finder, jury-rigging a handgrip from an L-bracket and bracing the camera against your face might buy you one more stop – down to 1/60s – but you will only have a sharp exposure perhaps one third of the time, and that’s a bit too uncertain for me. You can of course control exposure with ISO, NDs and the aperture ring; however there are times when you need to go slower; much slower. ISO 400 is the point at which I set the threshold for acceptable image quality; anything higher will still be usable, but not really up to what you’d expect for medium format. As you can see – f2.8 (the fastest lens for this system), 1/125s and ISO 400 is really not very dark at all.
At this point, we have to revert to using a tripod; while we’re at it, a cable release and locking up the mirror is a must. Fortunately, this is very easy – there’s a switch under the winding crank that takes care of it, making it very much second nature to operate. During my last trip to Tokyo, I carried one with me all the time – a very lightweight Gitzo GT1542 Traveller with an Arca-Swiss P0 mono ball head, made slightly lighter and more rigid by the removal of the centre column. I personally find tripod bags very inconvenient, so it was clipped into a belt holster and carried much like a sword of old. The weight was noticeable after a while, but not too bad; more inconvenient was me inadvertently whacking the tripod against things. I’d much rather have a between-the-shoulder-blades quiver, but I can’t seem to find one. The tripod worked well, with one caveat: it seems that Tokyo has a lot of areas that are actually structures built over manmade underground caverns, which means that street level is really the surface of a bridge. And cars driving over that bridge will cause vibrations in the surface, which are beyond the scope of the tripod’s ability to damp – we’re talking about oscillations you can feel through the soles of your shoes. Needless to say, there’s a very visible effect on image quality, and especially for long exposures. The only solution is to carefully time your shots between traffic. In crowded areas, not blocking thoroughfares and avoiding getting kicked are also a bit of a challenge.
I didn’t find power or storage to be an issue – despite the files being anywhere up to 70MB each. One 32GB card saw me through the entire week-long trip without deleting anything, and I though I carried three batteries – I always had one full one left at the end of the day, and sometimes a bit more.
What the shooting experience does force you to do is be very disciplined: there is nothing stopping you from carrying a lot of lenses (they actually aren’t that much bigger or heavier than 35mm FF lenses because the apertures are fairly modest), other than backache; however, I personally got the best results when I carried just one, or at most, two – an 80mm and a 150mm. I felt far less encumbered, for a start. Beyond that, the 80mm was my staple simply because it extended my shooting envelope by two stops – firstly, being f2.8 instead of f4, and secondly, being a shorter focal length allowing for slower shutter speeds before hand shake became apparent.
One thing I haven’t touched on is focusing, and alignment. We’ve got a few moving parts here: firstly, there’s mirror alignment to ensure that your flange-to-focusing screen distance is exactly the same as the flange-to-sensor distance; it isn’t always the case, especially on older bodies whose mount surfaces may well have worn down over the years. It is imperative to calibrate the mirror zero to ensure that what you see in the finder is exactly what you get on your sensor; this is a slow process that requires a little manual dexterity and a lot of patience.Having the correct focusing screen and viewfinder are very important too: the 45 degree prisms are useless for this task, in my opinion. Not only are they disorienting and impossible to use in the portrait orientation, they also don’t offer enough magnification.
Fortunately, the digital backs are supplied with a standard focusing screen that also has markings delineating the sensor area, and a square crop within that – this screen has an excellent matte, crispy micro prisms and a split image rangefinder in the centre. I find that together with the HC-4 prism, I have more than enough magnification and brightness to focus consistently: firstly, the combination gives 100% field coverage and 100% magnification with the 80mm lens; you can shoot with both eyes open. Secondly, eye relief is terrible because of the magnification, but it has an enormous +/-5 diopter adjustment range – enough even for my extreme myopia. As an added bonus, there’s the stability increase afforded by bracing the camera against your face when shooting it handheld. Even so, you have to remember that not only is depth of field razor thin, but the resolving power of the sensor in pixels per degree FOV is even higher than a D800E; combine this with the significantly higher mirror slap, and you’ve got a camera that’s extremely demanding to shoot handheld. My handheld hit rate is very, very low with this camera: much lower than film, lower than normal digital, and frankly, pretty abysmal. It’s not a compositional thing: it’s an image quality one; there’s no point in keeping a slightly motion-blurred or camera-shaken image, simply because I might as well have used something else to capture it with. The additional magic of medium format is gone. (On a tripod, this is a different story, of course – my hit rate is nearly 100%.)
However, the biggest reason why I put up with all of these foibles is a combination of rendering style, colour and ‘bite': the files are notionally 16 bit, which means a significantly wider gamut; the sensor has no anti-aliasing filter, either. The ability to discriminate between very close colours also translates into an ability to resolve higher-frequency and lower-contrast structures, which contributes to the impression of acuity and resolving power. Coupled with the nonlinear tonal response of CCDs – which coincidentally matches the impressions of human vision reasonably well – we’re left with a camera that produces extremely natural looking files; when you get everything right, almost zero post processing is required – it sees and replicates what you see; nothing more, nothing less. It is a different kind of transparency to the Otus/D800E combination: that still requires some tonal work for accuracy. The Hasselblad/CFV combination does not. By the same token, mess up the shot and you will not be shown any mercy: there is almost no latitude for repair afterwards. This level of clarity means that it’s much easier to convey the feeling of being there, or the illusion of reality, of course resulting in stronger images. MT
This set on flickr was shot entirely with the CFV-39.
Places left for 2014 Making Outstanding Images Workshops: Havana and London – click here for more information and to book!
Visit the Teaching Store to up your photographic game – including workshop and Photoshop Workflow videos and the customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved